io_loop error when running Dask Actors on SGECluster

What happened: Getting an error when trying to submit an Actor to a client started with an SGECluster cluster.

Using example code found here. Works fine with a LocalCluster . I get an io_loop error when running with an SGECluster.

Error:

AttributeError: type object 'Counter' has no attribute '_io_loop'

What you expected to happen: I expect the Dask Actor to work on an SGECluster

Minimal Complete Verifiable Example:

Example:

class Counter:
    """ A simple class to manage an incrementing counter """
    n = 0

    def __init__(self):
        self.n = 0

    def increment(self):
        self.n += 1
        return self.n

    def add(self, x):
        self.n += x
        return self.n

future = client.submit(Counter, actor=True)  # Create a Counter on a worker
counter = future.result()

counter.add(1)

Error:

AttributeError                            Traceback (most recent call last)
Input In [6], in <cell line: 1>()
----> 1 counter.add(1)

    211         return _OK(result["result"])
    212     return _Error(result["exception"])
--> 214 actor_future = ActorFuture(io_loop=self._io_loop)
    216 async def wait_then_set_result():
    217     actor_future._set_result(await run_actor_function_on_worker())

    187     else:
    188         return attr
--> 190 attr = getattr(self._cls, key)
    192 if callable(attr):
    194     @functools.wraps(attr)
    195     def func(*args, **kwargs):

AttributeError: type object 'Counter' has no attribute '_io_loop'

Anything else we need to know?:

Environment:

• Dask version: 2022.5.0
• Dask-jobqueue: 0.7.3
• Python version: 3.8.13
• Operating System: CentOs
• Install method (conda, pip, source): conda

[graingert]

@jacobtomlinson fixed

[mronda]

@graingert I think I fixed it! I had set set_as_default=False when starting my Client. I set that to True and now it does set Actor._client.io_loop . Thanks for the help ! You pointed me to some code that allowed me to debug that further